Proceedings of CLS 45(2), 103-116, 2009
© Chicago Linguistic Society 2009. All rights reserved.
Non-local binding in tenseless clauses
Sverre Stausland Johnsen
Harvard University
1 Introduction
Norwegian is commonly used in the literature to exemplify the binding domain of
reflexives. The descriptions and examples are taken from Hellan (1988, 1991),
who concludes with the claim summarized in (1):
(1) In Norwegian, the 3.person reflexive seg is freely bound out of non-finite
complement clauses, but cannot be bound out of a finite complement
clause (1988:73, 84, 1991:30f.).
(1) makes the condition of morphological finiteness decisive for the binding
domain of seg. This paper shows that another condition, tense, is a stronger factor
than finiteness, such that (1) is violated if the construction satisfies the condition
stated in (2) – the descriptive hypothesis in this paper:
(2) In Østfold Norwegian (ØN), the 3.person reflexive seg can be bound out
of a tenseless complement clause.
(2) makes no reference to the finite status of the complement clause, as it holds
for finite and non-finite clauses alike.
The fact in (2) follows as a consequence of the movement operations applied
independently for reflexive binding and restructuring of tenseless complement
clauses. As a result, (2) is analyzed as a restructuring effect in this paper, as stated
in (3), to be outlined in detail in section 5:
(3) Non-local binding of the 3.person reflexive seg in ØN is a restructuring
effect.
Given the differences in binding facts among the many Norwegian dialects
(Strahan 2003), the judgments in this paper, obtained through pairwise
comparisons, have been collected only from speakers from the state of Østfold,
located in the south-east corner of Norway.
2 Norwegian binding
Norwegian has a rather intricate system of pronominals, in that a pronoun is either
reflexive or non-reflexive, and simple or complex. All four possibilities exist, as
illustrated in (4):
104
SVERRE STAUSLAND JOHNSEN
(4)
Masculine 3.sg.
Bare Complex
Reflexive
seg
seg sjøl
Non-reflexive han han sjøl
The bare pronouns are furthermore possessive or non-possessive:
(5)
Masculine 3.sg.
Possessive Non-possessive
Reflexive
sin
seg
Non-reflexive
hans
han
For the reflexive binding cases discussed in this paper, the focus will be limited to
the non-possessive bare 3.sg./pl. seg, for the following reasons:
First of all, only the third person reflexive is morphologically distinct from its
non-reflexive counterpart (as seg-han above). In the 1./2.person, one form fills
both roles. Second, ØN conforms to Pica’s generalization, which states that
complex reflexives cannot be bound non-locally (Pica 1987:485, Stausland
Johnsen 2008:11). Since this paper treats non-local binding in ØN, the complex
reflexive seg sjøl falls out. Finally, although it is commonly assumed or explicitly
claimed that the possessive reflexive sin undergoes the same binding conditions
as seg (Hellan 1988:62, 74f., 1991:31, Faarlund et al. 1997:1164ff.), the only
published case where that assumption is tested shows that it does not (Strahan
2003:89). Given similar evidence in ØN (Stausland Johnsen 2007), and the fact
there is no reason to assume a priori that sin and seg behave identically, the
possessive sin is not treated in this paper. For the distribution of these pronominal
forms not discussed in this paper, see Hellan (1988:59ff.,) and Stausland Johnsen
(2008:3ff.)
‘Non-local binding’ as used here subsumes the terms ‘medium-distance
binding’ (binding out of a non-finite clause) and ‘long-distance binding’ (binding
out of a finite clause), as distinguished by Reuland & Koster (1991:23f.). Put in
traditional theoretical terms, non-local binding are instances of violation of
Chomsky’s binding condition A (Chomsky 1981:188).
There are many factors that will allow or inhibit non-local binding in
Norwegian, and some of them have been established in the literature already, such
as logophoricity (Moshagen & Trosterud 1990, Strahan 2001), animacy of the
local subject (Lødrup 2009), and factivity of the matrix verb (cf. Strahan
2003:89ff.). There are several factors influencing the possibility for non-local
binding in ØN as well, and this paper will isolate one of these factors: tense. The
role of tense does not seem to have been acknowledged as a factor in binding
previously in the literature – for any language.
NON-LOCAL BINDING IN TENSELESS CLAUSES
105
3 Østfold Norwegian (ØN) binding data
3.1 Finite complements
Most matrix verbs do not allow a non-locally bound seg in their finite
complements, exemplified below with the declarative sa ‘said’, the epistemic
trudde ‘believed’, and the psych-verb frykta ‘feared’:
(6a) *Reveni sa
[at noen
jakta
på segi]
(b) *Reveni trudde
[at noen
jakta
på segi]
(c) *Reveni frykta
[at noen
jakta
på segi]
The-fox said/believed/feared that someone chased on REFL
‘The fox said/believed/feared that someone was hunting him’
(7a) *Peri sa
[at noen
snakka om
segi]
Peter said that someone talked about REFL
‘Peter said that someone was talking about him’
(8a) *Peri sa
[at noen
la
et håndkle rundt
segi]
Peter said that someone laid a towel
around REFL
‘Peter said that someone put a towel around him’
(9a) *Hundeni trudde [at
noen
leika med segi]
The-dog believed that someone played with REFL
‘The dog believed that someone was playing with him’
When the matrix verb is a perception verb or the verb for ‘dream’,1 on the other
hand, a non-locally bound reflexive is allowed in its finite complement, as seen in
the equivalent examples of (6)-(9) below:2
(6d)
(e)
(f)
(g)
1
?Reveni
?Reveni
?Reveni
?Reveni
The-fox
hørte
[at noen
jakta
på segi]
så
[at noen
jakta
på segi]
lukta
[at noen
jakta
på segi]
drømte
[at noen
jakta
på segi]
heard/saw/
that someone chased on REFL
smelled/dreamed
‘The fox heard/saw/smelled/dreamed that someone was hunting him’
The verb for ‘dream’ often groups with perception verbs where the latter is a relevant
grammatical category for some phenomenon, as frequently encountered for evidentiality (cf.
Aikhenvald 2003:22, 2004:344ff.).
2
The examples have been marked with one question mark to render the fact that when asked
for judgments, speakers tend to indicate their preference for a complement with a non-reflexive
pronoun.
106
SVERRE STAUSLAND JOHNSEN
(7b) ?Peri hørte [at noen
snakka om
segi]
Peter heard that someone talked about REFL
‘Peter heard that someone was talking about him’
(8b) ?Peri kjente [at noen
la
et håndkle rundt
segi]
Peter felt
that someone laid a towel
around REFL
‘Peter felt that someone put a towel around him’
(9b) ?Hundeni drømte
[at noen
leika
med segi]
The-dog dreamed that someone played with REFL
‘The dog dreamed that someone was playing with him’
Given that the only overt difference between the constructions in the first and
second group of examples is in the matrix verb, it is tempting to link this with
previously observed effects of the matrix verb for non-local binding, namely the
matrix verb’s ability to license a logophoric or factive complement (see end of
section 2 above). Perception verbs, however, do not license logophoric pronouns
(Culy 1994:1061), and dream is not a factive (Simons 1997:1036), so these
explanations are not available here.3
3.2 Non-finite complements
In constructions with object control (OC) or ECM matrix verbs, some allow a
non-locally bound reflexive in their complements, and some do not, as seen in the
following examples:
(10a) *Lærereni
ba elevenej
[PROj stå
bak
segi] OC
The-teacher told the-students
stand (inf.) behind REFL
‘The teacher told the students to stand behind him’
(10b) ?Lærereni
lot elevenej
[PROj stå
bak
segi] OC
The-teacher let the-students
stand (inf.) behind REFL
‘The teacher let the students stand behind him’
(10c) Lærereni
så
[elevene
stå
bak
segi] ECM
The-teacher saw the-students stand (inf.) behind REFL
‘The teacher saw the students stand behind him’4
3
Reuland & Koster (1991:23f., Reuland 2006:96) make the strong claim that the only binding
that is possible out of finite clauses is logophoric binding. ØN binding as demonstrated in this
section is evidence to the contrary.
4
(10c) is considered better than (10b), but (10c) does not have the same clausal structure as
(10b). Whether the structure in (10c) contains an ECM-clause or a small clause is not crucial here,
as it under any approach is different from an OC construction. Since clausal structure has been
intensively investigated for Norwegian binding before (see (1)), I will not discuss this further here.
NON-LOCAL BINDING IN TENSELESS CLAUSES
107
(11a) *Lærereni beordra elevenej til [å PROj stå
bak
segi] OC
Theordered theto to
stand behind REFL
teacher
students
(inf.)
‘The teacher ordered the students to stand behind him’
(11b) ?Lærereni tvang elevenej til [å PROj stå
Theforced theto to
stand
teacher
students
(inf.)
‘The teacher forced the students to stand behind him’
bak
segi] OC
behind REFL
In the next section, we will see that the one thing that unifies the verbs that allow
non-locally bound reflexives in their complements – perception verbs and the
verbs drømme ‘dream’, la ‘let’, and tvinge ‘force’ – is that they select tenseless
complements.
4 Tenselessness
Tense and finiteness do not necessarily correlate. As a result, non-finite clauses
can be either tensed or tenseless (Stowell 1982:562f., Landau 2000:57, 2004:836,
Wurmbrand 2001:62ff., Wiklund 2007:38f.). Importantly, the same holds for
finite clauses – they can either be tensed or tenseless (cf. Varlokosta & Hornstein
1992:515f, Krapova 2001:117f., Landau 2004:831ff., Radišić 2006:9). As a
standard example of this distinction for non-finite clauses, cf. the difference
between (12a) and (12b) below:
(12a) John planned [to leave tomorrow].
(12b) *John tried [to leave tomorrow].
In (12a), the non-finite clause carries a future tense specification with respect to
the matrix clause, and it can therefore be modified with a temporal adverb
tomorrow. In (12b), on the other hand, the non-finite clause has no tense, and can
therefore not be modified by a temporal adverb that contradicts the tense of the
matrix clause.
4.1 Tenselessness in Østfold Norwegian
Wiklund (2007) demonstrates the tense properties of non-finite clauses in
Scandinavian. The evidence is of the kind in (12) above, and need not be repeated
here (see Wiklund 2007:38f.). The important conclusion is that the verbs be ‘tell’
and beordre ‘order’ have tensed complements (2007:48), whereas the verbs la
‘let’, se ‘see’, and tvinge ‘force’ have tenseless complements (2007:53, 56, 63).
108
SVERRE STAUSLAND JOHNSEN
As seen in section 3.2 above, the tenseless complements correlate with the
possibility of non-local binding.5
No evidence has been given in the literature of tenseless finite clauses in
Scandinavian, so the remainder of this section will present the evidence that the
finite complement clauses of perception verbs (and drømme ‘dream’) are tenseless
in ØN. In short, there are three domains in which this evidence manifests itself:
1) Sequence of tenses (SOT)
2) Double access reading (DAR)
3) Temporal adverb disagreement
The unique behavior of the complement clauses of perception verbs in these
domains supports the conclusion that they are tenseless, as will be seen in the
following.
4.1.1 Sequence of tenses (SOT)
In a past-under-past construction as (13), there are two available interpretations of
the subordinate past: simultaneous with the matrix tense (13a), or past-shifted
with respect to the matrix tense (13b) (Enç 1987:635):
(13) John said that Mary was pregnant
(13a) John said: “Mary is pregnant”
(13b) John said: “Mary was pregnant”
The simultaneous reading in (13a) is called an ‘SOT reading’. Just as in English,
ØN exhibits the same optionality between an SOT or past reading of a
subordinate past verb form:
(14) Per
sa
[at Kari var med
Peter said that Kate was with
(14a) Peter said: “Kate is pregnant”
(14b) Peter said: “Kate was pregnant”
barn]
child
SOT
past
When the matrix verb is a perception verb, on the other hand, the SOT reading of
the finite complement clause becomes obligatory:
(15) Per
så
[at Kari var med
Peter saw that Kate was with
(15a) Peter saw: /Kate is pregnant/
(15b) *Peter saw: /Kate was pregnant/
5
barn]
child
SOT
past
For a complete list of the relevant non-finite clauses in ØN and their binding properties, see
Stausland Johnsen 2008:45ff.
NON-LOCAL BINDING IN TENSELESS CLAUSES
109
(16) Per
drømte
[at Kari var med barn]
Peter dreamed that Kate was with child
(16a) Peter dreamed: /Kate is pregnant/
SOT
(16b) *Peter dreamed: /Kate was pregnant/
past
The simultaneous reading of the morphological past in SOT constructions is
standardly analyzed as the result of clausal tenselessness (Ogihara 1995:674, von
Stechow 1995:367, Kratzer 1998:101, Kusumoto 1999:82, Khomitsevich
2007:106). The obligatory SOT reading under perception verbs in ØN would
imply that perception verbs select for tenseless complements in ØN.6
The past morphology itself on the tenseless embedded verb has come about
through a copying or agreement operation with the matrix verb (Jespersen
1954:152, Ross 1967:333, 1986:198, Comrie 1985:114, Kusumoto 1999:64).
4.1.2 Double access reading (DAR)
A complement with present morphology can be embedded under a matrix past
verb, as in (17), giving DAR. This means that the embedded present is evaluated
with respect to both the matrix tense and the utterance time (Enç 1987:636f.). In
the case of (17), for the sentence to be true, Mary must be pregnant both at the
time of John’s utterance and at the time of the utterance of the sentence itself.
(17)
John said that Mary is pregnant
The same phenomenon occurs in ØN, as exemplified in (18):
(18) Per
sa
[at Kari er gravid]
Peter said that Kate is pregnant
‘Peter said that Kate is pregnant’
When the matrix verb is a perception verb in the past, however, a verb with
present morphology cannot be embedded under it:
6
The SOT reading in (14a) is also due to tenselessness in the complement. Unlike (15)-(16),
the SOT reading in (14) is optional. This is analogous with the interpretation of pronominals.
Regular pronouns are optionally free or bound, whereas reflexives must receive a bound reading.
The interpretational optionality of pronouns takes place at LF, whereas the obligatory bound
reading of reflexives is due to properties in narrow syntax (Reuland 2001a:440f.). In the same
vein, the optionality we see in (14) takes place at LF (Ogihara 1995:673ff., Abusch 1997:12ff.),
whereas the obligatory bound reading of the tense in (15)-(16) is due to the selectional restrictions
in narrow syntax. As a result, it is expected that operations in one domain of narrow syntax
(obligatory SOT) correlate with operations in another domain of narrow syntax (obligatory
binding). It is not expected that the outcome of an LF operation (optional SOT) has consequences
for operations in narrow syntax (obligatory binding). In sum, we do therefore not predict an
optional SOT reading of the sentences in (6)-(9) above to license non-local binding.
110
SVERRE STAUSLAND JOHNSEN
(19) *Per så
[at Kari er gravid]
Peter saw that Kate is pregnant
‘Peter saw that Kate is pregnant’
The ungrammaticality of (19) follows from the discussion of SOT in section 4.1.1
above. If perception verbs select for tenseless complements, they cannot select a
clause with present morphology embedded under past morphology, since the
present morphology in such a construction denotes temporal simultaneity (Enç
1987:642, Stowell 2007:446), which is not the same as tenselessness.
Furthermore, if the embedded clause is tenseless, it cannot acquire present
morphology through copy/agreement with the matrix verb, since the matrix verb
in this case has past morphology. (19) is therefore correctly ruled out in ØN as a
consequence of the analysis given for the obligatory SOT in (15)-(16).
4.1.3 Temporal adverb disagreement
The standard test for proving tenselessness in a complement clause is to let a
temporal adverb in the complement contradict a temporal adverb in the matrix
clause (Varlakosta & Hornstein 1992:516, Landau 2000:57, 2004:831ff., Krapova
2001:117, Wurmbrand 2001:74, Radišić 2006:9, Wiklund 2007:38f.). If the
complement has no tense, the adverb can be interpreted only if it takes matrix
scope. If there already is a temporal adverb taking matrix scope, however, and the
adverbs contradict each other, this disagreement cannot be reconciled. As a result,
an ungrammatical outcome of this test is a proof of complement tenselessness. An
example of this test for English can be seen in (20):
(20a) This morning, John planned to leave tomorrow tensed
(20b) This morning, John tried to leave (*yesterday)/(*tomorrow) tenseless
For finite complements, conflicts between temporal adverbs as in (20) are allowed
for all kinds of matrix verbs in ØN except perception verbs, as seen in (21):
(21a)
(b)
(c)
(d)
(e)
(f)
(g)
I dag
I dag
I dag
I dag
I dag
I dag
I dag
Today
sa
visste
huska
frykta
så
hørte
kjente
said/knew
remembered/feared
saw/heard/felt
Per
Per
Per
Per
Per
Per
Per
Peter
[at
[at
[at
[at
[at
[at
[at
that
det
det
det
det
det
det
det
it
regna
regna
regna
regna
regna
regna
regna
rained
i fjord]
i fjord]
i fjord]
i fjord]
(*i fjord)]
(*i fjord)]
(*i fjord)]
last year
The outcome of the temporal adverb disagreement test is therefore that the finite
complement clauses of matrix perception verbs are tenseless.
NON-LOCAL BINDING IN TENSELESS CLAUSES
111
4.2 Conclusion
The conclusions from all three tests in section 4.1 taken together provide ample
evidence that the finite complement clauses of perception verbs are tenseless in
ØN. Combined with the already available evidence that the non-finite
complement clauses of certain OC and ECM verbs such as la ‘let’, se ‘see’, and
tvinge ‘force’ are tenseless, we see that there is an evident correlation between the
tenselessness of a complement clause and its ability to license non-local binding.
For ØN, then, we can sum up the descriptive facts with the statement in (22):
(22)
Complement tenselessness licenses non-local binding
5 Analysis
5.1 Restructuring
‘Restructuring’ is a cover term for language specific phenomena which all share
the following characteristic, quoted from Roberts 1997:423:
(23)
“[...] processes and dependencies that are normally limited to a single
clause, can, where the higher predicate is of a particular type, take
place across clause boundaries”.
A recurring restructuring effect is that a syntactic item which is interpreted as the
internal argument of the embedded verb has an overt morphosyntactic realization
as the internal argument of the matrix verb. The classic cases of restructuring –
clitic climbing and long passives – both share this characteristic.
Two aspects of restructuring are important to emphasize here. One, a
necessary prerequisite for restructuring to take place is that the complement
clause is tenseless (Wurmbrand 2001:79ff., 91f., 2006:313, 321, Wiklund
2007:57f.). Two, although usually exemplified with non-finite clauses,
restructuring also takes place with finite clauses (cf. Guasti 1993:66ff., Progovac
1993:119, Terzi 1996:284ff., Stausland Johnsen 2008:24f.).
The restructuring effects themselves are caused by movement of the
embedded tenseless verb to the matrix verb, cf. Zushi (2001:33ff)., Wurmbrand
(2001:12f., 122ff., 2006:312f., 318ff.), Wiklund (2007:164), and the many
references therein.
5.2 Binding through movement
In the standard movement theory of binding, the reflexive moves covertly to the
T-projection, where it is in a local configuration with its subject binder in spec-TP
(Lebeaux 1983:726f., Chomsky 1986:175f., Pica 1987:490f., Battistella 1989:987,
Reinhart & Reuland 1991:302ff., Cole & Sung 1994:356, Cole et al.
2001:xxxviiiff., Reuland 2001a:455ff., Safir 2004:159ff.).
When an embedded reflexive is bound structurally (i.e. not logophorically) by
the subject in the matrix clause, it follows under this approach that the embedded
112
SVERRE STAUSLAND JOHNSEN
T must have undergone an independent movement to the matrix T, with piedpiping of the reflexive.
This is essentially the model that Reinhart & Reuland adopt in order to
account for the non-local binding examples in Hellan (1988), where the reflexive
seg in a non-finite complement clause is bound by the matrix subject
(1991:302ff., Reuland 2001b:355, Reuland 2006:97ff.):
verb raising with pied-piping
(24) Joni
bad
oss [snakke om
segi]7
reflexive moves to T
John asked us talk
about
‘John asked us to talk about him’
REFL
In (24), the embedded reflexive moves covertly to the T-projection in its clause.
Independently of that, the embedded T raises to the matrix T, making non-local
binding in these cases a by-product of pied-piping (Reuland 2006:97). As Reuland
acknowledges, the main problem with this analysis is that no motivation or trigger
for this verb movement is identified.
5.3 Non-local binding from restructuring
In ØN, tenseless complement clauses undergo restructuring, by which the
embedded T raises to the matrix T. If the embedded clause contains a reflexive, it
will be pied-piped to the matrix clause and bound by the matrix subject, as
proposed by Reinhart & Reuland. The lacking trigger for the interclausal verb
movement in Reinhart & Reuland’s model is therefore apparent in ØN –
tenselessness.
As a result, there is no need for any extra stipulations to account for the nonlocal binding cases addressed in this paper. Non-local binding can be treated as an
automatic consequence of independent processes taking place in ØN. Non-local
binding bears furthermore the hallmarks of a restructuring effect. In the same
fashion as clitic climbing and long passives, non-local binding is characterized by
having an internal argument of the embedded verb (the reflexive seg) realized as
if it were the internal argument of the matrix verb (i.e. as a locally bound reflexive
seg rather than a non-reflexive pronoun).
6 Restructuring in Østfold Norwegian
Restructuring is a process that manifests itself with tenseless complements in ØN
also outside of the non-local binding cases discussed in this paper. It is therefore
important to emphasize that restructuring is not posited solely to account for non-
7
Note, however, that (24) is ungrammatical in ØN.
NON-LOCAL BINDING IN TENSELESS CLAUSES
113
local binding. It exists independently of the binding facts. This section will
summarize the other restructuring effects in ØN.
In some Swedish and Norwegian dialects, among them ØN, the embedded
verb in non-finite tenseless clauses will enter a copy/agreement relation with the
matrix verb, and take on the tense morphology of the matrix verb, as illustrated in
(25)-(26):
(25) Ikke gidd
proi [å
Not bother (imp.)
to
‘Don’t bother to do it’
PROi
gjør
det]
do (imp.) it
(26) Jegi har ikke giddi
[å
I
have not bothered (perf.) to
‘I have not bothered to do it’
PROi
gjort
det]
done (perf.) it
The verb copying in (25)-(26) occurs as the embedded tenseless verb raises to the
matrix verb by restructuring (Wiklund 2007:164).8
6.1 Verb agreement & non-local binding
A prediction following from the facts of verb agreement and non-local binding in
non-finite clauses discussed above is that OC and ECM verbs which trigger verb
agreement in their complements will be the same verbs that license non-local
binding in their complements. The following examples show that this prediction
holds true:
la ‘let’
Verb copying
(27) Jeg hadde ikke latt
’ni [PROi gjort
det]
I
had
not let (perf.) him
done (perf.) it
‘I would not have let him do it’
Non-local binding
(28) ?Lærereni
lot elevenej
[PROj stå
bak
segi]
The-teacher let the-students
stand (inf.) behind REFL
‘The teacher let the students stand behind him’
8
When restructuring applies to tenseless finite complements, as discussed in section 4, the
verbs agree in their finite forms, i.e. pres.-pres. and pret.-pret. When restructuring applies to
tenseless non-finite complements, as discussed in this section, the verbs agree only in their nonfinite forms, i.e. inf.-inf., imp.-imp., and perf.-perf. Wiklund (2007:88) classifies the distinction
between finite and non-finite agreement as ‘full’ and ‘graded’ restructuring respectively. It lies
outside the scope of this paper to investigate the relationship between clausal structure and verb
agreement further.
114
SVERRE STAUSLAND JOHNSEN
se ‘see’
Verb copying
(29) Jeg har ikke sett
[’n gjort
det]
I
have not seen (perf.) him done (perf.) it
‘I have not seen him do it’
Non-local binding
(30) Lærereni
så
[elevene
stå
bak
segi]
The-teacher saw the-students stand (inf.) behind REFL
‘The teacher saw the students stand behind him’
be ‘tell’
Verb copying
(31) *Jeg hadde ikke bett
’ni [PROi gjort
det]
I
had
not told (perf.) him
done (perf.) it
‘I would not have told him to do it’
Non-local binding
(32) *Lærereni
ba elevenej
[PROj stå
bak
segi]
The-teacher told the-students
stand (inf.) behind REFL
‘The teacher told the students to stand behind him’
In the examples above, the matrix verbs la ‘let’ and se ‘see’ select tenseless
complements and allow both verb copying and non-local binding to take place.
The matrix verb be ‘tell’, on the other hand, selects a tensed complement, and as a
result, does not allow either verb copying or non-local binding.
7 Conclusion
This paper has provided evidence that in Østfold Norwegian, tenseless
complement clauses license non-local binding of reflexives. Tenseless
complements also trigger restructuring to take place. Within the movement theory
of restructuring and reflexive binding, the non-local binding cases in ØN fall out
as restructuring effects.
As a result, two syntactic phenomena in ØN that are seemingly quite unrelated
– interclausal verb agreement and non-local binding – can be seen as effects of
one single operation, namely T-to-T raising across clauses, or in short:
restructuring.
ØN seems to be the first language where tense, as opposed to finiteness, can
be demonstrated to restrict the binding domain. That this phenomenon has not
been reported for Norwegian before, however, gives promise that the same effect
is underreported or undiscovered in other languages, too. More research on
NON-LOCAL BINDING IN TENSELESS CLAUSES
115
similar ‘peripheral’ binding cases in this and other languages might reveal how
other, seemingly unrelated, aspects of the grammar might influence and restrict
(or widen) the binding domain of reflexives.
References
Abusch, D. 1997. Sequence of tense and temporal de re. L&P 20.1-50.
Aikhenvald, A. Y. 2003. Evidentiality in typological perspective. In Studies in Evidentiality, ed.
A. Y. Aikhenvald & R. M. W. Dixon, 1-31. Philadelphia: John Benjamins.
. 2004. Evidentiality. Oxford university press.
Battistella, E. 1989. Chinese reflexivization: a movement to INFL approach. Linguistics 27.9871012.
Chomsky, N. 1981. Lectures on Government and Binding. Cinnaminson: Foris publications.
. 1986. Knowledge of Language: Its Nature, Origin, and Use. New York: Praeger.
Cole, P., G. Hermon, & C.-T. J. Huang. 2001. Introduction. Long-distance reflexives: the state of
the art. In Syntax and Semantics 33, Long-distance Reflexives, ed. P. Cole, G. Hermon & C.T. J. Huang, xiii-xlv. Boston: Academic press.
. & L.-M. Sung. 1994. Head movement and long-distance reflexives. LI 25.355-406.
Comrie, B. 1985. Tense. Cambridge university press.
Culy, C. 1994. Aspects of logophoric marking. Linguistics 32.1055-1094.
Enç, M. 1987. Anchoring conditions for tense. LI 18.633-657.
Faarlund, J. T., S. Lie, & K. I. Vannebo. 1997. Norsk Referansegrammatikk. Oslo:
Universitetsforlaget.
Guasti, M. T. 1993. Causative and perception verbs. A comparative study. Torino: Rosenberg &
Sellier.
Hellan, L. 1988. Anaphora in Norwegian and the theory of grammar. Providence: Foris
publications.
. 1991. Containment and connectedness anaphors. In Long-distance anaphora, ed. J. Koster &
E. Reuland, 27-48. New York: Cambridge University press.
Jespersen, O. 1954. A Modern English Grammar on Historical Principles. Part IV – Syntax (Third
Volume). London: Bradford & Dickens.
Khomitsevich, O. 2007. Dependencies Across Phases. From Sequence of Tense to Restrictions on
Movement. Ph.D. dissertation, Utrecht University.
Krapova, I. 2001. Subjunctives in Bulgarian and Modern Greek. In Comparative Syntax of Balkan
Languages, ed. M. L. Rivero & A. Ralli, 105-126. Oxford University press.
Kratzer, A. 1998. More structural analogies between pronouns and tenses. In Proceedings from
Semantics and Linguistic Theory 8, ed. D. Strolovitch & A. Lawson, 92-110. Ithaca: Cornell
university.
Kusumoto, K. 1999. Tense in Embedded Contexts. PhD. dissertation. UMass Amherst.
Landau, I. 2000. Elements of Control. Structure and Meaning in Infinitival Constructions. Boston:
Kluwer academic publishers.
. 2004. The scale of finiteness and the calculus of control. NLLT 22.811-877.
Lebeaux, D. 1983. A distributional difference between reciprocals and reflexives. LI 14.723-730.
Lødrup, H. 2009. Animacy and long distance binding in Norwegian. NJL 32.111-136.
Moshagen, S. N. & T. Trosterud. 1990. Non-clause-bounded reflexives in Mainland Scandinavian.
WPSS 46.47-52.
Ogihara, T. 1995. The semantics of tense in embedded clauses. LI 26.663-679.
Pica, P. 1987. On the nature of the reflexivization cycle. In NELS 17, ed. J. McDonough & B.
Plunkett, 483-499. UMass Amherst.
116
SVERRE STAUSLAND JOHNSEN
Progovac, L. 1993. Locality and subjunctive-like complements in Serbo-Croatian. JSL 1.116-144.
Radišić, M. 2006. Subject raising and obligatory subject control. In Proceedings of the 2006
annual conference of the Canadian linguistic association, ed. C. Gurski & M. Radišić, 16 p.
http://ling.uwo.ca/publications/CLA2006/Radisic.pdf
Reinhart, T. & E. Reuland. 1991. Anaphors and logophors: an argument structure perspective. In
Long-distance Anaphora, ed. J. Koster & E. Reuland, 283-321. Cambridge University press.
Reuland, E. 2001a. Primitives of binding. LI 32.439-492.
. 2001b. Anaphors, logophors, and binding. In Syntax and Semantics 33, Long-distance
Reflexives, ed. P. Cole, G. Hermon & C.-T. J. Huang, 343-370. Boston: Academic press.
. 2006. Long-distance binding in Germanic languages. In The Blackwell Companion to Syntax
III, ed. M. Everaert & H. van Riemsdijk, 85-108. Malden: Blackwell publishing.
. & J. Koster. 1991. Long-distance anaphora: An overview. In Long-distance Anaphora, ed. J.
Koster & E. Reuland, 1-25. New York: Cambridge University press.
Roberts, I. 1997. Restructuring, head movement, and locality. LI 28.423-460.
Ross, J. R. 1967. Constraints on Variables in Syntax. PhD. dissertation. MIT.
. 1986. Infinite Syntax! Norwood: Ablex publishing corporation.
Simons, M. 2007. Observations on embedding verbs, evidentiality, and presupposition. Lingua
117.1034-1056.
Stausland Johnsen, S. 2007. Unpublished syntactic survey among 30 7th grade students in Askim,
Østfold.
. 2008. Tense and binding in perception verb complements. Ms., Harvard University.
von Stechow, A. 1995. On the proper treatment of tense. In Proceedings from Semantics and
Linguistic Theory 5, ed. M. Simons & T. Galloway, 362-386. Ithaca: Cornell university.
Stowell, T. 1982. The tense of infinitives. LI 13.561-570.
. 2007. The syntactic expression of tense. Lingua 117.437-463.
Strahan, T. E. 2001. The distribution of reflexive pronouns in Norwegian. AJL 21.159-169.
. 2003. Long-distance reflexives in Norwegian: a quantitative study. München: Lincom
Europa.
Terzi, A. 1996. Clitic climbing from finite clauses and tense raising. Probus 8.273-295.
Varlokosta, S. & N. Hornstein. 1992. Control in Modern Greek. In NELS 23, ed. A. J. Schafer,
507-521. UMass Amherst.
Wiklund, A.-L. 2007. The Syntax of Tenselessness. Tense/Mood/Aspect-agreeing Infinitivals. New
York: Mouton de Gruyter.
Wurmbrand, S. 2001. Infinitives. Restructuring and Clause Structure. New York: Mouton de
Gruyter.
. 2006. Verb clusters, verb raising, and restructuring. In The Blackwell Companion to Syntax
V, ed. M. Everaert & H. van Riemsdijk, 229-343. Malden: Blackwell publishing.
Zushi, M. 2001. Long-distance Dependencies. New York: Garland publishing.